Service level agreement

Availability and support commitments for enterprise plans. For evaluation and procurement; formal terms are in the signed agreement.

Last updated: March 2025

Uptime commitment

Target: 99.9% uptime per calendar month, measured at the application/API layer (excluding scheduled maintenance and exclusions below). Measurement is based on successful responses from the service health endpoint and core API paths from multiple regions.

SLA monitoring

Uptime is measured using the public GET /api/health endpoint and core API paths from multiple regions. Health checks run continuously; failures are alerted and counted toward monthly availability. Internal dashboards (e.g. admin health and observability) track error rates, latency, and dependency status. Status updates for incidents are communicated via support and, when applicable, a status page.

Exclusions

The following do not count toward downtime:

Scheduled maintenance (announced at least 72 hours in advance, or 24 hours for emergency)
Customer-caused issues (misconfiguration, abuse, or terms violation)
Force majeure (natural disaster, war, government action, etc.)
Third-party failures outside our reasonable control
Beta or non-production environments

Service credits (enterprise)

If we fail to meet the 99.9% uptime commitment in a calendar month, enterprise customers may request a service credit:

Uptime (month)	Credit
< 99.9% but ≥ 99.0%	10% of monthly fee
< 99.0% but ≥ 95.0%	25% of monthly fee
< 95.0%	50% of monthly fee

Credit is applied to the next invoice. Request in writing within 30 days of the end of the affected month. Credits are the sole remedy unless otherwise agreed in the customer contract.

Support response (enterprise)

Critical (P1): Production down or critical security — target initial response within 1 hour (business hours)
High (P2): Major feature impaired — within 4 business hours
Medium (P3): Minor feature or workaround available — within 1 business day
Low (P4): General questions — within 2 business days